Pulse Electrodeposition for Carbonate-Rich Deposits from Seawater

Abstract

Seawater electrodeposition is gaining renewed interest in the context of sustainable development, both to build climate-resilient coastal infrastructure and for ocean-based decarbonization applications. Most of the applications benefit from CaCO₃-rich deposits, but constant-voltage electrodeposition results in a mixture of CaCO₃ and Mg(OH)₂, especially at higher voltages where precipitation rates are more desirable. The use of pulse voltages can help control interfacial pH that dictates the precipitation reactions. Here, we explore the use of pulse electrodeposition as a function of pulse frequency and duty cycle to control deposit composition. The most CaCO₃-rich deposits were obtained under 10 Hz frequency and 10% duty cycle conditions for the voltage window investigated (−0.8 V to −1.2 V vs. SCE). While pulsing the voltage increases the amount of CaCO₃ deposited, the energy required per gram of CaCO₃ is significantly higher (14.5×) when compared to the base case of applying a constant voltage of −0.8 V vs. SCE. Further optimization of pulse conditions, electrode materials, and system configuration could lead to finding parameters that result in exclusively carbonate deposits without compromising precipitation rates, which may prove to be more useful for corrosion protection, coastal infrastructure, and other applications in sustainable development.

1. Introduction

Climate change disproportionately impacts coastal communities. With ever growing coastal populations [1] and the largest cities in the world being located in coastal zones, there is a pressing need to identify easily scalable strategies to protect coastal zones [2]. The most effective and commonly used climate interventions are engineered built infrastructure for coastal defense [3]. While such structures have become increasingly popular around the world [4], they result in changes to the local sediment composition and dynamics, eventually impacting the local habitat and ecosystem [5]. The use of materials sourced on-site and from the ocean would avoid such unintended impacts. Calcareous deposits that form on metal structures exposed to seawater is one such promising option [6].

Calcareous deposits are a result of electrochemical reactions leading to the precipitation of a mixture of CaCO₃ and Mg(OH)₂ from seawater. These deposits have been beneficially used to support and rejuvenate coral reefs [7,8], create sustainable concrete [9,10], prevent the corrosion of metal infrastructure [11,12,13], and more recently as a means to capture and remove CO₂ from seawater [14,15,16,17,18]. Though beneficial in these cases, calcareous deposits can prove detrimental to other electrochemical processes by causing scale formation on membranes, electrodes, and flowlines [19,20,21,22]. Due to the many applications and issues posed by calcareous deposits, the process has been widely researched and reviewed [12,23,24].

Prior studies on calcareous deposits have highlighted the sensitivity of the deposit composition and properties to various process conditions and the need for dedicated research into optimizing the process for each application. For example, to use calcareous deposits for coastal defense infrastructure, the mechanical strength of the precipitate is critical, which in turn is determined by the relative composition of CaCO₃ and Mg(OH)₂ in the precipitate. The Mg(OH)₂ is flaky and breaks off or undergoes dissolution, making it less ideal for structural applications [25], so CaCO₃-rich calcareous deposits are desirable.

In seawater, the major electrochemical reactions involved in the formation of calcareous deposits are the hydrolysis of water, the conversion of dissolved oxygen into hydroxyl ions, and the formation of chlorine gas with the generation of hydroxyl ions increasing local pH to induce the precipitation of CaCO₃ and Mg(OH)₂ as shown in Equations (1)–(5) below and in Figure 1a. Additional reactions can also occur in the system such as shifts in the bicarbonate/carbonate equilibrium and hypochlorite generation from the reaction of the generated Cl₂ and OH⁻; such side reactions minimize the rapid increase in the solution’s pH. Through a combination of electrochemical cell design and processing parameters, the side reactions can be accelerated or minimized, but such efforts are beyond the scope of this study. Other potential precipitation reactions such as the precipitation of MgCO₃ and Ca(OH)₂ are not observed during the electrochemical deposition due to slow reaction kinetics for MgCO₃ and greater solubility for Ca(OH)₂, requiring more extreme changes in solution pH before precipitating [26,27].

$${\mathrm{O}}_{2\left(\mathrm{g}\right)}+2{\mathrm{H}}_2\mathrm{O}+4{\mathrm{e}}^{-}\to 4{\mathrm{OH}}_{\left(\mathrm{aq}\right)}^{-}$$

(1)

$$2{\mathrm{H}}_2\mathrm{O}+2{\mathrm{e}}^{-}\to { \mathrm{H}}_{2\left(\mathrm{g}\right)}+2{\mathrm{OH}}_{\left(\mathrm{aq}\right)}^{-}$$

(2)

$$2{\mathrm{Cl}}_{\left(\mathrm{aq}\right)}^{-}\to {\mathrm{Cl}}_{2\left(\mathrm{g}\right)}+2{\mathrm{e}}^{-}$$

(3)

$$2{\mathrm{OH}}_{\left(\mathrm{aq}\right)}^{-}+{\mathrm{Mg}}_{\left(\mathrm{aq}\right)}^{2+}\leftrightarrow \mathrm{Mg}{(\mathrm{OH})}_{2\left(\mathrm{s}\right) }$$

(4)

$${\mathrm{OH}}_{\left(\mathrm{aq}\right)}^{-}+{\mathrm{HCO}}_{3\left(\mathrm{aq}\right)}^{-}+{\mathrm{Ca}}_{\left(\mathrm{aq}\right)}^{2+}\leftrightarrow {\mathrm{CaCO}}_{3\left(\mathrm{s}\right)}+{\mathrm{H}}_2\mathrm{O}$$

(5)

The precipitation reactions for Mg(OH)₂ and CaCO₃ occur at different pH values with reported pKsp values of 10.9 and 8.4, respectively, in seawater at 25 °C [23]. For Mg(OH)₂ specifically, the critical pH value to observe any precipitate was found to be 9.3 when testing in a seawater matrix free of CaCl₂ [28]. This difference in precipitation pH allows for different compositions of deposit to form based on the voltage or current applied to the system. When applying constant voltage, the chemical reactions near the electrodes can be controlled by selecting a voltage below the activation potential of undesired side reactions but sufficient for the activation of target reactions. In seawater, the main two reactions that need to be controlled are the oxygen reduction and hydrolysis reaction noted in Equations (1) and (2). The reduction of oxygen in water is known to have an activation potential around −0.45 V vs. SCE while the hydrolysis reaction has an activation potential around −1.05 V vs. SCE [29]. Maintaining a cathode potential below the activation of the hydrolysis reaction will prevent it from occurring while only moderately increasing local pH around the cathode through O₂ reduction. With the need for oxygen to diffuse to the cathode for this reaction to take place, the pH only slightly increases around the cathode as the reaction rate is limited, which favors the formation of CaCO₃. When applying a voltage that exceeds the activation potential for the hydrolysis reaction, no diffusion is necessary for the reaction, and the pH increases rapidly, allowing for Mg(OH)₂ to precipitate in large quantities along with evolution of hydrogen gas as noted in Equation (2). With bubbles forming at the same time as precipitation, the deposit generated will be porous and have cracks, which is not ideal for corrosion protection. In addition to differences in applied voltage and current profiles, differences in cation and anion concentration, temperature, electrode, and presence of organic molecules have all been shown to impact the composition and characteristics of the resulting deposits [24,28,29,30,31,32,33,34,35,36].

Though calcareous deposits have been well studied, investigation into how applying a pulsed-voltage profile influences the composition of the precipitates has received minimal attention. One of the potential benefits of using a pulsed voltage is the ability to have greater control over the composition and thickness of the film formed on to the electrode [37]. Given that the reaction to form CaCO₃ and Mg(OH)₂ is pH dependent, being able to control the local concentration of OH⁻ near the electrode using the pulsed field would allow control over the deposit composition. We illustrate the differences in local pH near the electrode under the static and pulsed potential scenarios as a function of distance from the cathode in Figure 1b. Under a pulsed potential, the generated OH⁻ ions from the pulse-induced electrochemical reactions are allowed to diffuse away from the cathode surface when the pulse is turned off, ensuring that the local pH increase is limited and does not exceed the critical pH for Mg(OH)₂ precipitation as shown by the pKsp threshold. In addition to controlling the local pH at the cathode, the pulse-off phase also allows for the diffusion of reactants to the cathode surface, permitting increased reaction rates during the subsequent pulse-on phase. Therefore, by manipulating the voltage pulses, the local pH can be controlled to result in electrodeposits rich in CaCO₃.

Some of the earliest investigations into the pulse electrodeposition of calcareous deposits were conducted by Zamanzade et al. [38,39,40]. In their experiments, they used a rotating disc electrode configuration and examined the influence of factors such as temperature, electrode rotation speed, Ca²⁺ and Mg²⁺ concentration, and the frequency of the pulses. Their work showed that by changing the frequency of the pulses, the resulting deposit composition would also be altered. It was shown that Mg(OH)₂ is favored at low frequency, while CaCO₃ in the form of aragonite is favored at high frequency [39]. More recently, Tu and Cottis have reported an optimum frequency for aragonite formation to be 100 Hz when applying a 50% duty cycle between current densities of 1 and 0 mA/cm² in the context of corrosion prevention [41]. While these studies demonstrate the feasibility of using pulse electrodeposition to control composition and obtain CaCO₃-rich materials, an accurate evaluation of Ca/Mg ratios in the deposits as a function of pulsing conditions is not reported but is critical to determining the feasibility of the approach for several applications.

In this study, a pulsed applied potential with varying frequencies and duty cycles is applied to a low-carbon steel cathode and inert anode to form a calcareous deposit and subsequently analyzed by inductively coupled plasma mass spectrometry (ICP-MS) to determine the concentration ratio of Ca/Mg in the deposit. This is the first investigation that uses analytical methods to track the composition of the deposit under different applied potential conditions to inform pulse-control strategies to maximize CaCO₃ content, which is desired for diverse applications.

2. Materials and Methods

Electrodeposition experiments were conducted using a 3-electrode cell with a 1 cm × 6 cm × 0.0635 cm low-carbon steel cathode (made from shim stock acquired from McMaster Carr), a Ag/AgCl reference electrode (leakless mini from eDAQ), and a Pt anode (Pt-coated Ti mini-electrode from eDAQ). To minimize the variability in electrode spacing from one experiment to the next, a custom-designed Teflon plate electrode holder was used (see Figure S1). The synthetic seawater electrolyte was prepared using Instant Ocean^® salt mix according to the package instructions (36 g of salt dissolved in 1 L of DI water). All experiments were conducted in duplicate under ambient conditions and lasted a minimum of 72 h to generate sufficient deposits for characterization in a solution volume large enough to experience only minimal changes in pH and concentrations over the course of the experiment. The voltage input anddata collection during the electrodeposition experiments was conducted using a VSP-300 potentiostat (BioLogic, Seyssinet-Pariset, France) and associated EC-lab software (V11.42). This study investigated constant voltage at −0.8 V and −1.2 V vs. SCE, based on the onset potentials for competing electrochemical side reactions (i.e., steel corrosion reaction >−0.65 V vs. SCE, O₂ reduction reaction <−0.45 V vs. SCE, and H₂O reduction reaction <−1.05 V vs. SCE). The pulsed-voltage experiments used the same voltage bounds (−0.8 V and −1.2 V vs. SCE) and pulse frequencies of 1, 10, and 100 Hz. The viable pulse frequency was determined through initial experiments that spanned a larger frequency range and did not show rapid degradation of the cathode (see Figure S2). The fraction of time the pulse is on is defined as a duty cycle, and experiments used 10, 30, and 50% duty cycles at a 10 Hz pulse frequency to investigate the influence of duty cycle on deposit composition.

All electrodes were carefully weighed before and after the experiment to obtain the deposit mass. The electrodes were dried in an oven overnight before taking post-experimental mass to account for any residual water. Mass losses due to cathode dissolution during the experiment were small due to the protective nature of the electrodeposit. To prepare the deposits for compositional analysis, the samples were washed with DI water to remove any salts that may have crystallized on the electrode surface upon electrolyte evaporation. The washed deposits were characterized by X-ray diffraction (XRD) using a powder X-ray diffractometer (Bruker D8 Discover Microfocus, Billerica, MA, USA) equipped with a Cu source. Scans were conducted using an accelerating voltage of 50 kV and current of 1000 mA and between 10 and 90° (2ϴ) at a step size of 0.02°. Simulated XRD patterns were generated using Mercury software (ver. 2024.2.0 (Build 415171), Cambridge Crystallography Data Centre) using relevant crystal structures obtained either from the Crystallographic Open Database or original sources [42,43]. To characterize sample morphology, scanning electron microscopy (SEM) imaging was used. Images were taken at 2 kV using an Apreo variable pressure SEM (Thermo Fisher Scientific, Waltham, USA). Energy dispersive X-ray spectroscopy (EDS) analysis was performed using a Phillips XL30 system to investigate the distribution of Ca/Mg in the sample.

To accurately measure the Ca and Mg composition of the deposits, ICP-MS analysis was performed using a Perkin Elmer NexION 2000 (Hopkinton, MA, USA) instrument. Samples were prepared by weighing deposits in a microanalytical balance and digesting in acid (70% by volume HNO₃) for a concentration of 5 g/L. The acid digestion was performed for over 48 h, and the resulting solutions were subjected to serial dilution so that the concentration of solutes was <200 μg/L and within the desired range for ICP-MS analysis. An internal scandium reference was used to calibrate the readings. The raw data were processed to obtain the Ca/Mg composition of the deposits.

3. Results and Discussion

To establish the impact of using a pulse voltage profile, baseline experiments were conducted at constant voltage, specifically at −0.8 V vs. SCE and −1.2 V vs. SCE, which were the limits of the voltage window used for the pulsed electrodeposition. The deposits formed on the low-carbon steel electrode (Figure 2a) were first examined by XRD (Figure 2b) to see how the different parameters of voltage and frequency impacted the crystallinity of the deposits. There were no observable changes in the XRD patterns of samples obtained from pulsed voltage compared with those from constant −0.8 V vs. SCE experiments. All samples were identified to contain majority aragonite-CaCO₃ (Figure 2b). Under a constant applied voltage of −1.2 V vs. SCE, the samples were primarily brucite, a polymorph of Mg(OH)₂, and the broad XRD peaks suggested poor sample crystallinity, in agreement with past reports on brucite electrodeposits [28,29,41]. Investigation of the morphology of constant-voltage electrodeposits using SEM showed mostly needle-like structures in aragonite-rich samples (Figure 2c) and platelet-like structures in brucite-rich samples (Figure 2d). The pulsed electrodeposits were aragonite-rich with needle-like morphology (see Figures S3–S5) similar to the deposit generated at −0.8 V vs. SCE. In addition to these crystal forms, there were dendrite and ribbon structures on the deposit as shown in Figure 2e. Based on further SEM and EDS analyses (Figures S6 and S7), the amorphous regions were identified as being primarily composed of Na in the case of dendritic structures and composed of Mg in the case of ribbon-like structures. The Na-rich regions were also Cl-rich, suggesting they were NaCl salt deposits that were produced upon electrolyte drying on the electrode surface. Such Na-rich deposits in SEM images were observed to be covering needle-like aragonite (Figure S7), confirming that they were surface deposits on top of the electrolytically grown structures. Similarly, there were other major elements of seawater observed in the samples (Figure S8). The presence of these salts is not correlated with electrodeposition conditions but rather the sample extraction and drying process.

Under pulsed electrodeposition vs. constant voltage experiments, we expect oscillations in pH near the electrode, resulting in different distributions and ratios of CaCO₃ and Mg(OH)₂ precipitation as the deposit grows farther away from the electrode surface. To investigate the spatial variability in Mg vs. Ca in the electrodeposits, we used EDS (Figure 2e, Figures S9 and S10). Our images clearly highlight the presence of distinct domains that are rich in Mg and Ca. Furthermore, while the Mg-rich regions do not show measurable Ca, the Ca-rich regions always contain low levels of Mg. This observation is consistent with the established understanding that CaCO₃ does not deposit on steel without a layer of brucite deposit on the electrode first [31,33,34]. This result suggests that completely eliminating Mg(OH)₂ from the deposit is not possible without further modification of the electrode surface to promote direct CaCO₃ deposition. Under pulsed conditions, the higher frequency pulses may allow a thin Mg(OH)₂ layer to deposit faster and more thinly before being covered by an initial layer of CaCO₃ in the pulse-off phase, allowing for Ca-rich deposits in the subsequent pulse-on phase.

To understand the rates of electrodeposition under the different conditions, the mass of the deposits was recorded for each experiment (

Table 1. Deposit mass after 72 h under constant and pulse potential profiles, with varying pulse frequency and duty cycle (DC).

Experiment	Trial 1 Mass (mg)	Trial 2 Mass (mg)	Mass Average (mg)	Energy per g CaCO3 (Wh/g)
Constant −0.8 V	17.2	21.7	19.5	1.53
1 Hz DC 10	18.9	24.2	21.6	6.77
10 Hz DC 10	21.3	16.9	19.1	22.2
100 Hz DC 10	24.4	22.6	23.5	28.4
Constant −1.2 V	160.9	190.3	175.6	676
10 Hz DC 30	23.5	27.0	25.3	53.1
10 Hz DC 50	69.4	49.6	59.5	55.2

). The data show an increase in deposit mass as a function of frequency when compared to a constant voltage of −0.8 V vs. SCE. The highest deposit mass was obtained under a constant voltage of −1.2 V vs. SCE, which was ~8× more than that obtained at −0.8 V vs. SCE. The observed trend in deposit mass is correlated with the highest applied voltage for each experiment. The higher voltages allow for water electrolysis reactions to occur, generating higher pH and leading to more precipitate deposition. However, most of the deposits obtained at −1.2 V vs. SCE are Mg(OH)₂, as the local pH escalates well above the pKsp for Mg(OH)₂ allowing for Mg-rich deposits to rapidly form on the electrode. By pulsing the voltage, this pH escalation is avoided, resulting in CaCO₃-rich electrodeposits. However, under pulsed conditions, a lower deposit mass is obtained. When the duty cycle for a given frequency of pulsing is increased, it results in applying higher voltages for longer periods of time, thereby increasing both the Mg content and the total mass of the deposits. By increasing the duty cycle from 10% to 30% and 50% for the 10 Hz condition, the deposited mass was observed to increase, with a significant increase in mass for 50% duty cycle, indicating that the pH may not be as well controlled, leading to significant Mg(OH)₂ precipitation.

To quantify the Ca/Mg ratio in the deposits, ICP-MS analysis was conducted, the results of which are shown in Figure 2f. The data show that the pulsed experiments increased the fraction of Ca-rich deposits over the baseline experiment conducted at −0.8 V vs. SCE. The Ca/Mg molar ratio reached a maximum for the 10 Hz experiments, with a molar ratio of 2.7, which is roughly 73 mol% Ca and 27 mol% Mg. The relatively high amounts of Mg measured by ICP-MS and the absence of brucite peaks in the XRD analysis of the pulse electrodeposition samples suggest the Mg compounds to be amorphous in nature. Further enhancing the frequency of the pulsed experiments to 100 Hz did not result in additional amplification of the Ca/Mg molar ratio, indicating that there is an optimized frequency that enhances the amount of CaCO₃ deposited for the given test conditions. The decrease in favorability for CaCO₃ deposition at higher frequency pulsing could be just a result of the system experiencing longer durations closer to −1.2 V vs. SCE from the slow kinetics of electrode polarity switching, combined with mass transport limitations that restrict the rapid diffusion of OH⁻ ions away from the electrode surface, allowing the local pH to linger above the critical pH value of 9.3 to induce more Mg(OH)₂ precipitation over CaCO₃. The experiment conducted at a constant voltage of −1.2 V vs. SCE almost exclusively deposited Mg (98.3 mol%) on the electrode. We note that in general the EDS results suggested higher Ca/Mg ratios than ICP-MS as the technique is surface-sensitive and does not fully factor in the Mg(OH)₂ layer on the electrode that is underneath the aragonite-CaCO₃.

Though the CaCO₃ content is shown to increase when a low duty cycle is used, the energy expenditure for the applied pulses adds significantly to the energy cost of carbonate mineralization. As shown in Figure 2g, the energy expenditure for applying a constant −0.8 V is significantly lower (0.022 Wh) than all the pulsed profiles considered here. The next lowest measured energy consumption value was 0.119 Wh from the 1 Hz 10% duty cycle case, showing a 5× increase in energy consumption with a 4.4× increase in energy used per gram of CaCO₃ generated (Table 1). The low energy consumption in the constant −0.8 V vs. SCE case is likely due to deposit layer becoming thick and electrically insulating, limiting the occurrence of key electrolysis reactions (Equations (1) and (2)). This causes the electrode to experience a low current and effectively reaches a saturation point where no more deposit is being formed without the expenditure of any energy. For the pulsed samples, the marginal gains in deposit mass when compared to the added energy expenditure suggests that the pulse-on phase drives forward the electrolysis reactions but does not generate a local pH high enough to induce larger gains in deposit mass. The energy consumption for the −1.2 V vs. SCE sample and larger mass suggests that the deposit formation at this voltage is not as limited, when it comes to generating a sufficient local pH to drive precipitation. The deposit from this constant voltage case is primarily brucite, but applying a larger duty cycle in the pulsed cases could lead to more deposit mass with a stronger preference for CaCO₃ formation.

Based on the fact that the most Ca-rich deposits were obtained with 10 Hz pulse frequency, subsequent experiments were conducted at this frequency, with higher duty cycles (30% and 50%). ICP-MS characterization and total energy consumption of the resulting electrodeposits are shown in Figure 3a–c. From the XRD data, all samples were identified to be aragonite-CaCO₃ with no variability in the degree of sample crystallinity. While XRD showed no presence of Mg(OH)₂, ICP-MS analysis of the digested deposits highlights the presence of considerable amounts of Mg in all samples. The Mg content of the deposits increased from 27 mol% in 10% duty cycle to 54.0 mol% and 69.5 mol% in 30% and 50% duty cycle experiments, respectively. This indicates that either the Mg deposits are amorphous or poorly crystallized. The SEM image of the 10 Hz duty cycle 50% (Figure S11) shows a significant amount of amorphous material in between aragonite pillars, further supporting that the Mg deposits may not be detected by XRD. Though the total deposit mass increased as expected, the selectivity for CaCO₃ is lost when higher duty cycles are used. It may be possible to slowly ramp up the duty cycle to increase the deposit mass without compromising the selectivity for CaCO₃, but the overall energy usage, as shown in Figure 3c, increases with duty cycle, making it a less attractive approach for processes trying to minimize their energy or CO₂ footprint. Thicker and denser deposits with CaCO₃ can be worth it in the context of corrosion prevention as a thicker layer can help prevent the corrosion of underlying metal structures, especially for structures deployed within seawater.

Our results show that pulsed voltage with shorter duty cycles offers control over the interfacial pH to maximize precipitation of CaCO₃. Given that the CaCO₃ deposits are denser and less prone to dissolution/reprecipitation than Mg(OH)₂, the pulse voltage electrodeposits have more desirable mechanical properties for coastal construction and anti-corrosion applications. Further, pulse voltage could also be used for the reduction in Mg(OH)₂ deposition on electrodes used in ocean-based electrochemical processes to suppress scaling-related inefficiencies. For example, Kim et al. recently demonstrated an electrochemical marine CO₂ removal technique using Ag/AgCl and Bi/BiOCl electrodes [22], where a key factor limiting process performance was the formation of Mg(OH)₂ precipitates on the electrodes. Voltage pulsing in such systems could reduce some of the observed precipitation.

4. Conclusions

In this work, the impact of using a pulsed voltage on the composition of calcareous electrodeposits was investigated. Generally, CaCO₃ and Mg(OH)₂ are co-precipitated, and maximizing the CaCO₃ content in the deposits offers enhanced mechanical strength, which is relevant for coastal structures and for carbon mineralization applications. The reduced Mg(OH)₂ precipitation could also be of value to minimizing scale formation on electrodes in marine electrochemical systems. Pulsed voltage controls the interfacial pH at the electrode which dictates the precipitation reactions. We conducted experiments under constant and pulsed-voltage conditions, with pulsed experiments varying the pulsing frequencies and duty cycles. The electrodeposits were characterized for their composition, morphology, and Ca/Mg content to determine experimental conditions that would result in CaCO₃-rich deposits. Lower duty cycles and frequencies resulted in more Ca-rich precipitates, while higher frequencies resulted in more deposit mass. Our data show factors such as voltage boundaries, pulse frequency, and duty cycles to impact the deposit composition and deposition rates. We demonstrate the utility of pulse electrodeposition for the formation of calcareous deposits, but dedicated optimization will be needed to achieve the target deposit composition for a given application and to obtain faster deposition rates for the process to be relevant for practical use. Further, cost–benefit analyses of using pulse electrodeposition will need to factor project siting, sources of electricity, etc., when it comes to determining its efficacy for sustainability-focused end uses.